In lhjohn/WaltersDementiaModelJson: A Package Skeleton for Replicating Published Prediction Models

```r library(PatientLevelPrediction) knitr::opts_chunk$set( cache=FALSE, comment = "#>", error = FALSE, tidy = FALSE)

# Introduction

This vignette describes how one can use the function 'createAgeCovariateSettings' to define age covariates. This is useful if you want to include age covariates other than age in years or 5-year age groups.

You will need to create a function that takes age in years as input and maps it.

## createAgeCovariateSettings

This function contains the settings required to define an age covariate where you can specify any mapping of age.  

The setting 'ageMap' specifies a mapping of age to use.  For example, if you wanted to create a variable log(age)^2 then:

```r

function(x){return(log(x)^2)}

data <- data.frame(Input = c('covariateName', 
                             'ageMap',
                             'covariateId',
                             'analysisId'),
                   Description = c('The name of the covariate',
                                   'The function for mapping age in years',
                                   'The id of the covariate - generally ageId*1000+analysisId',
                                   'The analysis id for the covariate'
                                   ) )
library(knitr)
kable(data, caption = 'The inputs into the create function')

Example

To add the variable natural logaritm of age squared:

ageCov1 <- createCohortCovariateSettings(covariateName = 'log age squared',
                                            ageMap = function(x){return(log(x)^2)}
                                            analysisId = 459,
                                            covariateId = 1*1000+459)

To add the variable age squared:

ageCov2 <- createCohortCovariateSettings(covariateName = 'log age squared',
                                            ageMap = function(x){return(x^2)}
                                            analysisId = 459,
                                            covariateId = 2*1000+459)

To include both covariates you can create a list of these settings:

ageCov <- list(ageCov1, ageCov2)

lhjohn/WaltersDementiaModelJson documentation built on March 3, 2021, 3:33 p.m.